Journal of Cold Regions Engineering

The unfrozen water content is a result of soil freezing, and its variation is closely related to temperature. To study the change in unfrozen water content, a relationship between water pressure and ice pressure is built based on the Clapeyron equation. Then, using an assumption of coincident faces between the water–ice interface and the water–air interface, a formula is deduced to estimate the unfrozen water content via the soil-water potential curve combined with the capillary pressure theory. In this process, the soil particle size distribution curve is introduced to determine the soil pore-size distribution and obtain the soil-water potential curve. Finally, verification and analysis are performed by comparing these results with the pre-existing test data for unfrozen water content. The results show that unfrozen water content can be calculated by means of the soil particle size distribution curve, and the simulation accuracy has a positive correlation with the degree of saturation occupied by water. Moreover, the equivalent aperture size increases as the mean grain size of the soil increases, and the equivalent aperture size increases sharply and approaches the mean particle size when the mean is within a range of relatively large values. Additionally, the value of scale parameter α has an obvious influence on the unfrozen water content curve, as enlargement of α causes an increase in the unfrozen water content and makes water freezing in soils more difficult, and vice versa.

Both cone penetration test (CPT) tip resistance data and ground-penetrating radar (GPR) amplitude data vary at the interface of snow layers of different density. Therefore, relationships between these data should be observable, enabling spatial extrapolation of CPT tip resistance values using GPR. Quantitative analysis of GPR amplitude data occurs in pavement analysis, but these techniques have not been applied to snow. GPR sounding using a commercially available ground-coupled 400 MHz antenna was conducted in the immediate vicinity of numerous CPT holes in Antarctic firn. Comparison between CPT tip resistance and GPR amplitude data reveals that extrapolation of point CPT resistance data is possible over large spatial areas by tracking GPR horizons that equate with CPT resistance value. In addition, GPR amplitude and polarity can reveal information about relative snowpack density. GPR can be used efficaciously with CPT snow resistance data, enabling efficient extrapolation of snow physical properties across large areas. Complementary use of GPR with CPT can enhance site investigation procedures for the development of polar infrastructure.

In this paper, salt expansion tests were conducted on coarse-grained sulfate soils to investigate salt expansion characteristics. The experimental results indicate that the formation and growth of salt crystals are influenced by environmental temperature and humidity; during the cooling process, the specimen expands due to the growth of salt and ice crystals. The heave in the positive temperature range increases with increasing sulfate content, while the heave in the negative temperature range reaches a peak value when the salt content varies between 0.4% and 0.7%, indicating that the combined effect of precipitation and freezing contributes to significant volumetric expansion. The initial precipitation temperature increases with increasing pore solution concentration before it reaches 30°C. The freezing point is influenced by soil particles, pore solution concentration, and initial water content. The freezing point first decreases with increasing salt content and tends to be stable afterward when the water content is relatively high. However, the freezing point exhibits a continuous decreasing tendency if salt crystal precipitation occurs before freezing. An empirical model is proposed to predict the salt heave ratio in the positive temperature range based on salt crystal volume changes, and the relationship between the precipitation and freezing processes in the negative temperature range is examined.

Roads and other infrastructure in Western Alaska are difficult to design and build owing to the high cost of construction materials and the ice-rich permafrost foundation. The present study explores methods to treat locally available frozen materials for satisfactory engineering applications. Stabilizing agents including cement, perlite, nanosilica, and polypropylene fiber (PP-fiber) were used at various mix ratios to treat pulverized frozen silty sand. Subsequently, a suite of laboratory tests was performed on thawed and/or cured specimens to assess the thaw strain, shear-strength parameters, elastic wave velocities, and thermal conductivity of the treated soil. Scanning electron microscopy images were obtained to reveal the microstructure of the stabilized frozen soil. It was found that the addition of cement, perlite, nanosilica, and PP-fiber greatly reduces the thaw strain and improves the shear strength of the stabilized soil. A combination of cement and perlite at 3% each can be very effective and economical in reducing the thaw strain, enhancing the shear strength, and decreasing the thermal conductivity, and is recommended for field trial.

Hoarfrost on roadways and bridges can cause slippery and dangerous conditions for motorists. To reduce the costs and environmental impact of countermeasures, road authorities wish to optimize their winter maintenance operations. To support this, good knowledge of the hoarfrost formation process is required. This paper presents a laboratory setup designed and built to study hoarfrost formation in detail under controlled conditions. The accumulation of hoarfrost and the stability of the main controlling parameters (air temperature, surface temperature, and relative humidity) are quantified. By using an open-loop wind tunnel with warm, humid air flowing over a cold stone surface, we produced conditions similar to those of frost formation on a road with good stability. The hoarfrost growth rates were found to be within the range of field measurements published previously. The growth rates were constant during each test and were directly related to the driving force created by the difference in the water vapor pressure in the air and at the stone surface.

Damage to flexible pavements is known to be greatly influenced by climatic factors. Among them, partial thawing conditions encountered during winter warming periods, as well as in the early period of spring thaw, are reported as critical conditions for flexible pavement structures in cold regions. Experimental research was conducted using a laboratory test pit and a heavy vehicle simulator in order to document flexible pavement response during these critical conditions. Freezing of a typical experimental pavement was induced at −10°C. Thawing at 20°C was initiated afterwards from the surface downward. Pavement response and conditions were monitored using a thermistor string, as well as stress and strain sensors in all the pavement layers. The pavement response was measured regularly throughout thawing. It was more specifically documented for asphalt concrete temperatures of 10°C, 0°C, and −5°C when thaw depths reached 300, 600, 900, and 1,200 mm. The paper documents the pavement response for various partial thaw and temperature conditions. Weakening was found to be moderate in the pavement structural layers but reached a maximum for the subgrade soil once the pavement was completely thawed.

The mountainous regions of the southern United States experience substantial snow loads. This study develops a data-driven method for determining 50-year ground snow load (GSL) in high-elevation, low-latitude regions and applies it to Coconino County, Arizona. A new relationship between snow depth and GSL is developed, as applying a relationship developed for the entire United States would result in an underestimation of GSL. The peaks-over-threshold method provided a reliable alternative to the commonly used annual maxima method to determine GSL at stations with short or incomplete data records. Simple kriging with locally varying means was used to spatially model GSL in the region, which is strongly controlled by elevation with variation around topographic features. The effect of elevation was not accurately represented in the current ground snow loads prescribed by Coconino County, leading to values that are too high for lower-elevation areas and too low for higher-elevation areas.

Deep ice lenses or excess ice in permafrost in Arctic oil fields on Alaska's North Slope, the Mackenzie Delta, or Eastern Siberia in Russia create challenging issues for oil well completion design. As thaw subsidence adjacent to oil wells deepens, it induces large drag loads through negative skin friction and leads to strain damage and even failure of the well casing. This paper presents a case study of the surface casing failure mechanism resulting from the subsidence of thawing permafrost by elastoplastic finite-element (FE) modeling. The study site, permafrost thaw evaluation results, and field observations of surface casing damage are summarized. Details of the FE model, material and interface modeling considerations and parameters, and load application steps are described. Modeling results include ground subsidence; stress and strain redistribution in the thawed permafrost near the casing; the shear stress at the surface casing–soil interface; and the vertical stress, strain, and deformation of the casing string. The model limitations are also discussed. Results show that the most likely failure mechanism of the surface casing string is plastic lateral buckling. Surface casing subsidence, possible failure mode, and locations predicted by this model compare favorably with field observation data.

To better understand the interactions between ships and ice, analyze ice–ice collisions, and provide parameters for the design of ships and offshore structures, it is important to study the interaction between ice and waves. Understanding these interactions is crucial to the analysis of collision velocity and impact during ship–ice interactions. In this study, experiments were carried out focusing on factors that influence sea ice motion using a motion tracking system in a towing tank. Two conditions, a fixed wave height and a fixed wave steepness, were considered in examining the motion of synthetic ice models with various shapes, lengths, and thicknesses; analyzing changes in the surge, the heave amplitudes, and the drift velocity; and calculating the kinematic energy involved. The results demonstrate that the shape, length, and thickness of ice blocks, as well as the water wavelength, affect surge and heave. The wavelength and ice thickness were found to be the most important factors. Characterization of the kinematic energy acquired by the ice model was also carried out.

To investigate the stress–strain behavior of frozen soils, a program of triaxial compression tests was designed and carried out on samples of unfrozen and frozen cohesive (CL) and granular (SP) soils and pure ice. The experiments involved study of the influence of freezing, temperature reduction, and loading rate on the stress–strain characteristics of the frozen ground. The aim of this study is to assess the possibility of using the artificial ground freezing (AGF) technique in the excavation and tunneling in Line 2 of the Tabriz Subway project. The results show that freezing of the CL soil has no significant effect on the type of soil behavior (strain hardening), whereas freezing of the SP soil changes its strain-hardening behavior to strain softening. The effect of freezing on the increase in shear strength of the saturated SP soil is much greater than that of the saturated CL soil; however, the rate of increase in the shear strength due to freezing and temperature reduction is much greater for the CL soil. Freezing and reduction in temperature cause an increase in the elastic modulus of all the materials tested in the present study. In addition, the shear strength and elastic modulus of these materials increase with loading rate.